French firms deploy agentic AI on NVIDIA, from orbit to public offices

Companies in France are putting agentic computing to work in space, financial services, the public sector and scientific research, according to NVIDIA. The company says the French ecosystem has moved "from ambition to application", with open models and NVIDIA infrastructure at the centre.
- Bull will extend European AI server production to NVIDIA Vera Rubin Ultra NVL72 systems, planned for 2028, with rack integration in Angers.
- Altair-1, the first satellite of a 50-satellite, $1 billion constellation, reached orbit with NVIDIA Jetson AGX Orin modules running Mistral's models.
- Twake.ai, built on NVIDIA Nemotron, is the sole referenced sovereign digital workplace for more than 2,500 French public organizations.
- Mirakl has onboarded more than 70 million products and cut seller onboarding to under two hours.
- A Côte-d'Or digital-twin agent managing nearly 5,800 kilometers of roads and buildings has cut operating costs by 10%.
Sovereign infrastructure, from factory floor to orbit
NVIDIA says French groundwork began with the AI Action Summit in February 2025 and the Choose France Summits in 2025 and 2026, which anchored major compute commitments on French soil.
Bull is extending its European AI server production to NVIDIA Vera Rubin Ultra NVL72 systems, planned for 2028. Trays will be produced in Europe, with rack integration at its Angers factory, which was inaugurated on Friday, Oct. 2.
In space, Altair-1 is the first satellite in a $1 billion, 50-satellite constellation with on-board autonomous capabilities. It launched this month and has successfully reached orbit, according to NVIDIA. Loft Orbital integrated it, and Orbitworks, a joint venture between Loft Orbital and Marlan Space, built it in Abu Dhabi. The satellite flies NVIDIA Jetson AGX Orin edge modules and runs Mistral's models for real-time Earth observation and autonomous operations in orbit.
Open models for sovereign data
NVIDIA says its Nemotron open models are in production at French organizations that require sovereign, trustworthy data.
- LINAGORA and Docaposte secured the CANUT framework agreement. Twake.ai, built on NVIDIA Nemotron, becomes the sole referenced sovereign digital workplace for more than 2,500 French public organizations.
- SBS is launching a sovereign platform, in collaboration with NVIDIA, that lets banks deploy governed generative and agentic services within their own environments.
- Mistral introduced the Mistral Large 4 open model under private preview last week. NVIDIA describes it as offering state-of-the-art, open weight performance in cyber defense, manufacturing, finance and multimodal tasks.
- Fondation Inria, with NVIDIA's support, is launching a working group on open source advanced modeling and tool adoption. Findings are expected in the spring of next year.
Agents at work across industries
NVIDIA lists a series of customer examples in retail, public services, telecommunications and science.
- L'Oréal is moving cosmetic formulation discovery from bench chemistry to in-silico screening with the open source NVIDIA ALCHEMI Toolkit, running on its own GPU cluster. It is targeting discovery up to 100x faster.
- Mirakl has onboarded more than 70 million products with fine-tuned and open models on NVIDIA GPUs and inference software, cutting seller catalog onboarding to under two hours.
- Omi combines approved 3D digital twins with open models optimized on NVIDIA for 3x faster inference, helping brands place accurate products in AI-generated marketing visuals.
- ThinkDeep and Akila built a digital-twin agent for Côte-d'Or using NVIDIA Cosmos open world models, Nemotron and Omniverse libraries. It manages nearly 5,800 kilometers of roads and buildings and has cut total operating costs by 10%, with a 30% target as deployment progresses.
- Port Sud de France runs goods processing and port security on ThinkDeep's DeepBrain platform, powered by Nemotron and Cosmos.
- Enchanted Tools, a Paris-based robotics company, is deploying social robots to support patients in healthcare settings, including pediatric cancer care.
- Bouygues Telecom and Prisme.ai launched AI Studio V2 on NVIDIA full-stack infrastructure, so employees can build their own agents in a controlled environment.
Photoroom and Vestiaire Collective are also cited for the measurable business impact of advanced visual modeling on marketplace performance.
Quantum and what comes next
In quantum computing, Alice & Bob has integrated the NVIDIA NVQLink platform for its superconducting cat qubits, showing how GPUs can make quantum error correction more efficient. Quandela is using NVQLink to build an integration between a DGX Spark system and its photonic qubits, exploring hybrid quantum-classical applications.
The examples are presented by NVIDIA, and the figures, such as cost reductions and speedups, are company-reported or stated as targets. They show a French ecosystem where hardware, open models and sector applications are being combined, from public administration to banking and satellites. NVIDIA founder and CEO Jensen Huang's GTC Berlin keynote is scheduled for Wednesday, Oct. 21, at 11 a.m. CEST.


